Skip to main content

​
Intent-Driven Retrieval

When you search your memory, CORE doesn’t just keyword match. It first understands what kind of question you’re asking, then routes to the optimal search strategy. Different queries need different approaches - asking for your preferences is fundamentally different from asking what happened last week.

​
The 5 Query Types

​
1. Aspect Query

You’re asking about a specific type of information.
  • “What are my coding preferences?”
  • “Show me my decisions about authentication”
  • “What are my directives for code review?”
CORE extracts the relevant aspects (Preference, Decision, Directive), filters statements by those aspects, and returns structured facts.

​
2. Entity Lookup

You’re asking about a specific person, project, technology, or concept.
  • “Tell me about Sarah”
  • “What do I know about TypeScript?”
  • “Show me everything about the CORE project”
CORE finds the entity node in the graph and returns all statements where that entity is the subject or object, grouped by aspect.

​
3. Temporal Query

You’re asking about what happened in a specific time range.
  • “What happened last week?”
  • “Show me my work from January”
  • “What did we decide yesterday?”
CORE extracts the time range from natural language and filters episodes and statements by when they were recorded or when events occurred.

​
4. Exploratory

You’re asking for broad context or a catch-up.
  • “Catch me up on recent work”
  • “What have I been working on?”
  • “Show me my recent activity”
CORE returns recent session summaries grouped by topic - broad discovery mode with no specific filtering.

​
5. Relationship Query

You’re asking about how two entities are connected.
  • “How do I know Sarah?”
  • “What’s the connection between TypeScript and CORE?”
  • “How is the payment service related to Stripe?”
CORE finds statements connecting both entities and can compute shortest paths between indirectly connected entities through the graph.